Mass Spectrometric Analyses of
Post ‐Translationally Modified Proteins
Dissertation
For the Award of the Degree
“Doctor of Philosophy” (Ph.D.)
Division of Mathematics and Natural Sciences of the Georg‐August‐Universität Göttingen
Submitted by He‐Hsuan Hsiao From Taipei, Taiwan
Göttingen 2010
Members of Thesis Committee Dr. Henning Urlaub (Instructor)
Bioanalytical Mass Spectrometry Group Max Planck Institute for Biophysical Chemistry Prof. Dr. Ralf Ficner (Reviewer)
Department of Molecular Structural Biology Georg-August-University Göttingen
Prof. Dr. Frauke Melchior (Reviewer) ZMBH
University of Heidelberg
Date of Oral Examination: August 9, 2010
Affidavit
I hereby declare that this dissertation “Mass Spectrometric Analyses of Post-Translationally Modified Proteins” has been written independently and without unauthorized assistance. This dissertation has not been submitted elsewhere for any academic award or qualification.
He-Hsuan Hsiao July, 2010
Göttingen, Germany
Acknowledgement
First of all, I would like to express my deepest thankfulness to Dr. Henning Urlaub for the opportunity to accomplish my PhD in his laboratory and for funding and supervising this work. Thanks for thinking for me and working wholeheartedly on my manuscripts for publication among his busiest time. He is an greatly kind man and provides many chances to his students for being engaged in their scientific researches.
I thank Prof. Dr. Ralf Ficner of the University of Göttingen and Prof. Dr. Frauke Melchior of University of Heidelberg for serving on my thesis committee, and giving the suggestion of my work.
I am grateful to my colleagues - Uwe Pleßmann, Monika Raabe, Johanna Lehne, Mads Grønborg, Florian Richter, Carla Schmidt, Miroslav Nikolov, Katharina Kramer, Ilian Atanassov, Romina Hofele and Ling Yun. Your expertises in biology, mass spectrometry and bioinformatics have broaden my horizon. It is a joy working with all of you.
Thanks to my collaborators - Prof. Dr. Reinhard Lührmann and the people in his laboratory who have directly contributed to the phosphoproteomic project in spliceosome; Prof. Dr. Jürgen Wienands and Thomas Oellerich on the SLP65 project;
Prof. Dr. Markus Wahl and Xiao Luo on the NusB-S10-RNA interaction project;
Prof. Dr. Frauke Melchior, Dr. Erik Meulmeester and Benedikt Frank on the SUMO project.
Finally, I would especially like to thank my parent, brothers, girlfriend Chia-Yan Wu and all my friends in Taiwan for being the stone of support that I have built my life upon. I share my success to all of the wonderful people who have supported, assisted, and loved me during my PhD study at University of Göttingen. Without these support, I would not have achieved so much.
Table of Contents
Members of Thesis Committee……….……….I
Declaration……….…II
Acknowledgement………...III
Table of Contents……….…IV
List of Abbreviations……….……...…VIII
Summary……….……1
Chapter 1 - General Introduction………..………..………..………3
1.1 Proteomics………...3
1.2 Proteomic Analysis Techniques………...…….4
1.3 Separation Technology………..4
1.3.1 Sodium dodecyl sulfate polyacrylamide gel electrophoresis………...…5
1.3.2 Two-dimensional gel electrophoresis………..5
1.3.3 Chromatography………6
1.3.4 Affinity-Based Enrichment for Sub-Proteome………6
1.4 Mass Spectrometry……….7
1.4.1 Ion Sources……….8
1.4.2 Mass Analyzer………9
1.4.2.1 Quadrupole Mass Analyzer………...10
1.4.2.2 Ion Trap Mass Analyzer……….…10
1.4.2.3 Time-Of-Flight Mass Analyzer………...11
1.4.2.4 Fourier-Transform Ion-Cyclotron Resonance Mass Analyzer…..……11
1.4.2.5 Orbitrap Mass Analyzer………..…11
1.4.3 Tandem Mass Spectrometry………..12
1.5 Data Analysis……….13
Chapter 2 - A High-Throughput Method for Phosphopeptide Enrichment of Spliceosomal Proteins and Its Application………..………15
2.1 Summary………15
2.2 Introduction………15
2.3 Experiment Sections………19
2.3.1 Materials………19
2.3.2 A Crude Mixture of Nuclear snRNP Particles, Individual U snRNP, Spliceosomal Complexes and SR Proteins Purification………19
2.3.3 Ethanol Precipitation………...………19
2.3.4 In-Solution Digestion………...19
2.3.5 In-Gel Digestion………20
2.3.6 In-House TiO2 Microspin Column Fabrication……….20
2.3.7 Comparison of Different TiO2 Enrichment Procedure………20
2.3.8 NanoLC-ESI and -MALDI Mass Spectrometry Analysis………21
2.3.9 Interpretation of Tandem Mass Spectra………...22
2.4 Result and Discussion……….22
2.4.1 TiO2 Microspin Column………...22
2.4.2 Sensitivity of the TiO2 Microspin Column……….23
2.4.3 Optimizing Phosphopeptide Enrichment with TiO2 microspin column for In-Solution Digestion……....……….25
2.4.4 Global Profiling of Phosphopeptides from A Crude Mixture of Nuclear snRNP Particles, Individual U snRNP, Spliceosomal Complexes and SR Proteins……….29
2.4.5 Novel and Known Kinase Motifs in Spliceosomal Proteins………...34
2.4.6 Application of TiO2 Microspin Column………..36
2.5. Conclusion………40
Chapter 3 - Efficient Enrichment of Intact Phosphoproteins prior to Mass Spectrometric Analysis……….42
3.1 Summary………..……….42
3.2 Introduction………42
3.3 Experiment Sections………44
3.3.1 Materials………44
3.3.2 U1 snRNPs Purification………..44
3.3.3 Ethanol Precipitation………...44
3.3.4 Acetone Precipitation………..44
3.3.5 Optimized Calcium Phosphate Precipitation………...44
3.3.6 In-Solution Digestion………..……….45
3.3.7 TiO2 Enrichment Procedure………...45
3.3.8 Mass Spectrometry Analysis………..45
3.3.9 MASCOT Database Searching.,………46
3.4 Result and Discussion……….46
3.4.1 Comparison of Efficiency and Specificity among Different Alkaline Earth Metal ………...………..46
3.4.2 Optimal Condition for Phosphoprotein Precipitation………..48
3.4.3 Effect of Different Denatured Reagents………50
3.4.4 Examine the Power of our Phosphoprotein Isolation Method………...53 3.4.5. Identification of Phosphorylation Sites from U1 Small Nuclear
Ribonucleoproteins………55
3.5. Conclusion………58
Chapter 4 - Pseudo-Neutral-Loss Scan for Selective Detection of Phosphopeptides and N-Glycopeptides using Liquid Chromatography Coupled with a Hybrid Linear Ion-Trap / Orbitrap Mass Spectrometer……….………..59
4.1 Summary………59
4.2 Introduction………59
4.3 Experiment Sections………62
4.3.1 Materials………62
4.3.2 In-Solution Digestion………...62
4.3.3 Mass Spectrometry………..62
4.4 Result and Discussion……….64
4.4.1 The Principle of Pseudo-Neutral-Loss………..64
4.4.2 Identification of Phosphopeptide by Pseudo-Neutral-Loss Scan……….65
4.4.3 Identification of N-Glycopeptide by Pseudo-Neutral-Loss Scan………..69
4.5 Conclusion……….71
Chapter 5 - “ChopNSpice”, a Mass Spectrometric Approach That Allows Identification of Endogenous Small Ubiquitin-like Modifier-conjugated Peptides………...…..73
5.1 Summary………73
5.2 Introduction………73
5.2.1 SUMOylation………73
5.2.2 Challenges for Identification of SUMOylation by MS-based Approach……….……75
5.3 Experiment Sections………76
5.3.1 Materials………76
5.3.2 Software………77
5.3.3 In vitro SUMOylation Assays………..77
5.3.4 Cell culture, Immunoprecipitation and Immunoblotting………..77
5.3.5 Protein Digestion………..77
5.3.6 Liquid Chromatography and Mass Spectrometry………78
5.3.7 Data Analysis………78
5.4 Result and Discussion……….79
5.4.1 ChopNspice………..79
5.4.2 Identification of SUMO Conjugation Sites in vitro………...84
5.4.3 Increasing Sensitivity by using “High Mass” Acquisition………....86
5.4.4 Identification of SUMO-Conjugated Sites in vivo………87
5.5 Conclusion……….89
References………..……….91
Appendixes………103 Appendix 1. Phosphorylation sites identified from spliceosomal proteins by using
in-house TiO2 microspin column enrichment………..………..…103 Appendix 2. MS and MS/MS spectra of phosphopeptides derived from human PRP6 and PRP31………….………120 Appendix 3. Phosphorylation sites identified from U1 snRNPs by using CPP method in combination with in-house TiO2 microspin column enrichment……….………….124 Appendix 4A. MASCOT searching result against swissprot bovine database by using regular data-dependent acquisition………127 Appendix 4B. MASCOT searching result against swissprot bovine database by using pseudo-neutral-loss acquisition………..131 Curriculum Vitae………..……….…132
List of Abbreviations
2DE, two-dimensional gel electrophoresis 3D ion trap, three dimensional ions trap AC, alternate current
ACN, acetonitrile BCR, B-cell receptor CAA, chloroacetamide
CHCA, -cyano-4-hydroxy-cinnamic acid CID, collision-induced dissociation CPP, calcium phosphate precipitation DC, direct current
DDA, data-dependent acquisition DeoxyHex, deoxyhexose
DHB, 2,5-dihydroxybenzoic acid DTT, dithiothreitol
ESI, electrospray ionization FA, formic acid
FT-ICR, fourier-transform ion-cyclotron resonance Glu-Fib, [Glu]-Fibrinopeptide B
HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid Hex, hexose
HexNAc, N-acetylhexosamine IAA, Iodoacetamide
IMAC, immobilized metal ion affinity chromatography kDa, kilodalton
LC, liquid chromatography m/z, mass-to-charge ratio
MALDI, matrix-assisted laser desorption / ionization-time of flight mass spectrometer μg, microgram
μL, microliter mg, milligram min, minute
MOAC, metal oxide affinity chromatography MS, mass spectrometry
MS/MS, tandem mass spectrometry
MudPIT, multi-dimensional protein identification technology nL, nanoliter
NeuAc, N-acetylneuraminic acid
NeuGc, N-glycolylneuraminic acid NusB, N utilization substance protein B p53, cellular tumor antigen p53
PA, phthalic acid
PMF, peptide mass fingerprinting ppm, parts per million
pre-mRNA, precursor messenger ribonucleic acid PRP31, pre-mRNA-processing factor 31
PRP6, pre-mRNA-processing factor 6 PTMs, post-translational modifications pSer, phosphoserine
pThr, phosphothreonine pTyr, phosphotyrosine
Q-TOF, quadrupole-time-of-flight
RanGAP1, Ran GTPase-activating protein 1 RF, radio frequency
RNA, ribonucleic acid RP, reverse phase rpm, rounds per minute S10, 30S ribosomal protein S10 SAC, strong anion exchange SCX, strong cation exchange SDS, sodium dodecyl sulphate
SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis SLP-65, B-cell linker protein
snRNPs, small nuclear ribonucleoproteins Sp100, nuclear autoantigen Sp-100
SR proteins, serine/arginine-rich proteins SUMO, small ubiquitin-like modifier TFA, trifluoroacetic acid
TiO2, titanium dioxide TOF, time-of-flight
Uba2, SUMO-activating enzyme subunit 2 USP25, ubiquitin carboxyl-terminal hydrolase 25 UV, ultraviolet
Summary
Protein post-translational modifications (PTMs) possess key functions in the regulation of various cellular processes. In this thesis, several new technologies are developed to map protein PTMs, containing phosphorylation, glycosylation and SUMOylation.
Five major topics are presented in this thesis. In Chapter 1- General Introduction, describes what is proteomics, the importance of protein PTMs, the main techniques in proteomic analysis, including the principle of technologies for protein and peptide separation, the theory of mass spectrometry (MS) and concept of proteomic data analysis.
In Chapter 2 - A High-Throughput Method for Phosphopeptide Enrichment of Spliceosomal Proteins and Its Application, a disposable TiO2 microspin column is fabricated in-house for enrichment of phosphopeptides. The method offers several advantages, including high-throughput, easy to use, low cost, high selectivity and sensitivity. In combination with different proteomic strategies, 1381 unique phosphorylation sites corresponding to 390 distinct proteins were identified in spliceosomal proteins. We further applied this method to explore the phosphorylation sites on PRP6, PRP31 and SLP65 and to study protein-RNA interaction by UV-induced crosslinking reaction.
In Chapter 3 - Efficient Enrichment of Intact Phosphoproteins prior to Mass Spectrometric Analysis, a straightforward and reliable phosphoprotein purification procedure was developed based on calcium phosphate precipitation (CPP). Integration of TiO2 microspin column, a total of 192 unique phosphorylation sites corresponding to 45 distinct proteins were identified from the U1 small nuclear ribonucleoproteins (snRNPs); of these, 59 phosphorylation sites were not reported previously.
In Chapter 4 - Pseudo-Neutral-Loss Scan for Selective Detection of Phosphopeptides and N-Glycopeptides using Liquid Chromatography Coupled with a Hybrid Linear Ion-Trap / Orbitrap Mass Spectrometer, a pseudo-neutral-loss scan on a hybrid LTQ-Orbitrap MS was built up for selectively probing phosphopeptides and glycopeptides. The presence of known characteristic mass pair (phosphoric acid or monosaccharide residues) in the spectrum during in-source collision-induced dissociation (CID) was selected to trigger MS/MS and multi-stage activation MS3 fragmentation. Our method is compatible with nano-liquid chromatography (nano-LC) for separation of complex peptide mixtures without any further enrichment. The consequent spectra provide peptide sequence identification and modified site assignment as well as information of the glycan structure.
In Chapter 5 - “ChopNSpice”, a Mass Spectrometric Approach That Allows Identification of Endogenous Small Ubiquitin-like Modifier-conjugated Peptides, a novel, user-friendly and straightforward database search tool was developed, called “ChopNSpice”, to unambiguously determine the mammalian SUMO1 and SUMO2/3 conjugation sites in vitro and in vivo by mass spectrometry in combination with MS-based search engines like MASCOT or Sequest. High mass data dependent acquisition (DDA) is highly suitable for the accurate detection and sequencing of larger peptides and additionally facilitates detection of lower abundant SUMO-conjugates. We demonstrated the power of ChopNSpice software in combination with proteomic strategy, resulting in the identification of 10 SUMOylated proteins corresponding to 17 distinct sites in endogenous Hela-S3 cells. 15 SUMOylated sites were identified in this study appeared to be novel, which may provide a valuable resource to the biological research community.
Chapter 1 - General Introduction
1.1 ProteomicsThe Human Genome Project and its sister projects for other organisms mark the culmination of twentieth-century biology and have been a tremendous success, rapidly building up a new scope of scientific landscape for the century [1, 2]. Following in the footsteps of genomics, the next step has been the development of proteomics. The term “proteome” was coined by Marc Wilkins in 1995 to describe the total set of proteins encoded by a genome. The word
“proteome” was defined as “the PROTEin complement expressed by a genOME” [3, 4].
Proteomics is the study of proteome and was first coined by Peter James in 1997 to make an analogy with genomics [5]. Proteins, the main carriers of biological activity. The function of protein depends on the precise amino acid sequence, the modifications, the protein concentration, the association with other proteins, and the extracellular environment.
Accordingly, the proteomics is concerned with determining protein structure, modifications, protein expression levels, protein-protein interactions, localization, and cellular roles of as many proteins as possible.
Proteins are converted to their mature forms through a complicated succession of post-translational processing and decoration events, namely post-translational modifications, PTMs. Far from being merely decoration, PTMs of proteins control many biological processes, and examining their diversity and individual functions are critical for understanding mechanisms of cell regulation. As many as 300 PTMs of proteins are known to occur physiologically (http://www.unimod.org/). PTMs are covalent processing events that change the properties of a protein by proteolytic cleavage or by addition of a modifying group to one or more amino acids. Many of the PTMs are regulatory and reversible, most notably are protein acetylation, methylation, phosphorylation, ubiquitination, SUMOylation, and etc., which control biological function through a multitude of mechanisms. It is known that many enzymes and receptors are switched "on" or "off" by phosphorylation and dephosphorylation, such as the enzyme, glycogen synthase kinase-3 (GSK-3), which is phosphorylated by protein kinase B (PKB) as part of the insulin signaling pathway [6]. Ubiquitin is added to proteins as a tag that predestines them for proteolytic degradation [7]. Small ubiquitin-like modifier (SUMO) is found to be attached on many eukaryotic nuclear proteins which functions in the regulation of nuclear-cytosolic transport, transcriptional regulation, apoptosis, protein stability, response to stress, and progression through the cell cycle [8].
Despite the great importance of PTMs for biological functions, large scale studies have been hampered by lacking suitable methods. Mass spectrometry (MS) is currently the most versatile technology to directly determine PTMs due to its sensitivity and selectivity. However, the identification of PTMs still remains a substantial challenge owing to the low stoichiometry of
modified proteins in combination with large amounts of unmodified proteins in biological sample, which interfere the detection in MS. Hence, direct analysis of PTMs requires isolation of the correctly processed proteins in a significant amount followed by MS-based proteomics, we believe this will lead to a great contribution for the study of protein PTMs.
1.2 Proteomic Analysis Techniques
The extraordinary achievements of current proteomics are based largely on the successful developments in the fields of “separation technology”, “mass spectrometry” and “data analysis”.
Once joined, the three disciplines provided a powerful tool to study the proteomics shown in Figure 1.1.
1.3 Separation Technology
Mass spectrometry-based proteomics is highly dependent and tightly linked to separation technologies that simplify incredibly complex biological samples prior to MS analysis would be crucial. The detection of low abundance species is required front-end separation due to the overshadowed signal of high abundance species. To reduce the complexity, proteins or
Figure 1.1. Mass spectrometry-based proteomic strategy. The proteins are pre-fractionation, typically separated by SDS-PAGE. The gel lane is cut into several slices, which are then degraded enzymatically into peptides, where C-terminus are protonated amino acids (arginine or lysine) when trypsin is used, providing an advantage in subsequent peptide sequencing. The peptides are separated by one or more steps of chromatographic technologies. The eluted peptides are ionized by either ESI or MALDI. Finally, the resulting mass spectra are searched against protein database to obtain the information of peptide sequences.
peptides can be resolved into fractions by using various separation methods, including electrophoretic techniques (SDS-PAGE, 2D gel electrophoresis), multi-dimensional chromatography (size exclusion, ion exchange, reverse phase chromatography) and affinity purification.
1.3.1 Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) SDS-PAGE is a technique widely used in the separation of proteins according to their electrophoretic mobility (molecular weight) [9, 10]. The proteins have identical charge per unit mass due to the binding of SDS. Since the charge density is constant, the rate of migration depends on the resistive frictional force, thus small proteins migrate faster than big proteins.
The distance traveled in a fixed time period is a log function of the molecular weight.
1.3.2 Two-dimensional gel electrophoresis
From the mid 1970s, proteomics was pursued with two-dimensional gel electrophoresis (2DE).
The proteins in a sample are separated by isoelectric point and protein mass Each observed protein spot is quantified by its staining intensity. Selected spots are excised, digested and analyzed by mass spectrometer for protein identification shown in Figure 1.2.
2DE has been a mature technique for more than 35 years [11] and was the first technique capable of supporting the concurrent quantitative analysis of large numbers of gene product
Figure 1.2. Protein identification by two-dimensional gel electrophoresis (2DE) coupling with mass spectrometric peptide mass fingerprinting (PMF). The biological sample is first separated by isoelectric point, called isoelectric focusing (IEF). In the second dimensional separation, an electric potential is again applied, but at a 90 degree angle from the first field
,
separating the proteins on the basis of their molecular weight. The protein of interest is digested by adequate enzyme, and then analyzed by mass spectrometer. The resulting masses of the peptides of the unknown protein are then compared to the theoretical peptide masses of each protein in the protein database for protein identification, called peptide mass fingerprinting.[12, 13]. Peptide mass fingerprinting (PMF) coupled with matrix-assisted laser desorption / ionization-time of flight mass spectrometer (MALDI-TOF MS) has become highly efficient at the identification of 2DE separated proteins [14-18]. In this method, the unknown protein of interest is first enzymatically digested into peptides, where masses can be accurately measured with a mass spectrometer. These masses are then compared to the theoretical peptide masses of each protein calculated from a known protein database. The results are statistically analyzed to find the best match. For PMF, the matching of four peptides representing 10 % of the sequence does not constitute a reliable hit and should not be listed as a positive identification. In contrast, six peptides representing 20 % of the sequence may be adequate for a tentative identification. There are several shortcomings for this approach. 2DE has limited dynamic range. The protein sequence has to be present in the database of interest.
The presence of a mixture can significantly complicate the analysis and potentially compromise the results. Therefore, protein identification based solely on PMF should no longer be acceptable and must be complemented by tandem mass spectrometry (MS/MS) to achieve sufficient specificity of identification.
1.3.3 Chromatography
Gel-based technologies have been traditionally used with off-line MALDI analysis. In contract, the multi-dimensional chromatography is usually directly coupled to on-line electrospray ionization (ESI) analysis, a continuous separation, due to their buffers compatibility with ESI.
Two major chromatographic materials are widely used for separating peptide mixtures, a reverse phase (RP) and a strong cation exchange (SCX). The reverse phase material separates proteins or peptides based on their hydrophobicity, given high resolution, efficiency, reproducibility. However, the single dimension of separation might not provide sufficient peak capacity to separate peptide mixtures as complex as those generated by the proteolysis of protein mixtures, for example, total cell lysates. Another material, SCX, is integrated with reverse phase as the part of a two-dimensional chromatography, improving the resolving power of separation based on peptide charges and hydrophobicity interaction. This technique is known as multi-dimensional protein identification technology (MudPIT) [19, 20]. High complexity sample is first loaded onto an SCX, and is eluted in a series of salt concentration steps. Each eluted fraction is loaded onto an RP column either off-line or directly eluted into an ESI mass spectrometer. The MudPIT analysis is subdivided the sample into several independent MS runs, which increases the confidence of protein identification and the dynamic range of the measurement.
1.3.4 Affinity-Based Enrichment for Sub-Proteome
Another important separation technology is affinity purification which is often used to isolate proteins with PTMs based on a highly specific biological interaction such as that between
antigen and antibody, enzyme and substrate, or receptor and ligand. The low stoichiometry of PTMs such as phosphorylation, glycosylation and acetylation requires enrichment steps prior to MS analysis. Antibody-based affinity enrichment is widely used for the detection of PTMs in proteins. The antibody purification approach has been successfully used for global analysis of protein lysine acetylation [21], arginine methylation [22], tyrosine phosphorylation [23] and so on. In addtion, the N-glycosylated proteins can be isolated by lectins [24] and the protein containing phosphotyrosine can also be purified by protein containing Scr homology 2 (SH2) domains [25]. However, highly specific antibodies, receptor and substrate are not always available for PTMs of interest. The other affinity enrichment method is based on chemical derivation of the modifying group that derives a “tag” for affinity purification. For example, the O-phosphorylated residues [26] and O-GlcNAc residues [27] (serine or threonine) can be modified an affinity tag by a beta-elimination / Michael addition reaction. Special issues should be taken care of due to the possible loss of the low abundant peptides with PTMs and false-positive protein identification of side-products of chemical reactions
1.4 Mass Spectrometry
By fundamental definition, mass spectrometry is designed to measure the mass-to-charge ratio (m/z) of gas phase ions. It consists of an ion source that converts analyte molecules into gas phase ions, a mass analyzer that separates the m/z of the ionized analytes, and a detector that records the number of ions at each m/z value, as shown in Figure 1.3.
Biological mass spectrometry, the technological base of current proteomics studies, was first catapulted to mainstream prominence with the development of ion sources, the electrospray ionization (ESI) [28] and matrix-assisted laser desorption ionization (MALDI) [29-31]
techniques. The most notable the discovery and development of protein ionization methods are recognized by the 2002 Nobel Prize in Chemistry (http://nobelprize.org/). Widespread Figure 1.3. Main components of a typical mass spectrometry. Ion source for ion generation, mass analyzer for ion separation and detector to transform analogue signals into digital signals and record a mass spectrum.
mass analyzers are quadrupole, ion trap, time-of-flight (TOF), Fourier-transform ion-cyclotron resonance (FT-ICR) and Orbitrap which can be coupled with either ESI or MALDI ion source.
1.4.1 Ion Sources
The development of electrospary ionization (ESI) and matrix-assisted laser desorption ionization (MALDI), the two soft ionization techniques capable of ionizing peptides and proteins, revolutionized protein analysis using MS as shown in Figure 1.4.
The ESI source produces ions from solution at atmospheric pressure shown in Figure 1.4A.
ESI is driven by high voltage applied between the end of the capillary column and the inlet of the mass spectrometer. The processes of ESI involve creation of electrically charged spray, Taylor cone [32, 33], followed by formation of an aerosol of charged droplets and desolvation of analyte-solvent droplets. Eventually, ions become free of the solvent that surrounds them, and these ions make their way by voltage-driven into the mass analyzer of the mass spectrometer. The ESI ion formation are the multiply charged species and sensitivity to the analyze concentration and flow rate. Multiply charged ions enable mass spectrometers with limited m/z ranges to analyze higher molecular weight molecule. ESI ionizes the analytes out of a solution and is therefore readily coupled to liquid-based (for example, chromatographic and electrophoretic) separation tools.
Unlike ESI, the MALDI source produces ions from solid phase shown in Figure 1.4B. MALDI relies on the utilization of a matrix compound capable of absorbing ultraviolet (UV) light. The matrix and sample are mixed in the appropriate solvent and deposited onto a sample plate.
The solvent is evaporated, forming co-crystallized analyte-matrix molecules. MALDI Figure 1.4. The common ionization sources for proteomic research. (A) The scheme of the electrospray ionization (ESI) process. (B) The scheme of matrix-assisted laser desorption ionization (MALDI) process.
sublimates and ionizes the samples out of a dry, crystalline matrix via UV laser pulses. The MALDI matrix absorbs laser energy and transfers the energy to the analyte, whereas the rapid laser heating causes desorption of matrix and analyte ions into gas phase [31, 34]. MALDI is normally used to analyze relatively simple samples, whereas ESI integrated with liquid chromatography (LC) is preferred for the analysis of complex samples. Although MALDI can still be coupled to LC, the effluent from LC run must be deposited on a sample plate and mixed with the MALDI matrix, a process that has thus far proven difficult to automate [35].
1.4.2 Mass Analyzer
The mass analyzer is, literally and figuratively, central to mass spectrometry. For proteomics research, five basic types of mass analyzers are commonly used: quadrupole, ion trap, time-of-flight (TOF), Fourier-transform ion-cyclotron resonance (FT-ICR) and Orbitrap mass analyzer shown in Figure 1.5. They are very different in design and performance, each with its own strength and weakness. These analyzers can be stand alone or, in some cases, put together in tandem to take advantage of strengths of each. In the context of proteomics, key parameters of mass analyzer are sensitivity, resolution, mass accuracy and the ability to generate information-rich ion mass spectra from peptide fragment (tandem mass or MS/MS spectra).
Figure 1.5. Mass spectrometers used in proteomic research. (A) Quadrupole mass spectrometer, the ions are separated by time varying electric fields between four rods, permitting a stable trajectory only for the ions of a particular desired m/z. (B) Three-dimensional ion trap mass spectrometer, the ions maintain a stable trajectories inside the device as a result of the application of a radio frequency voltage to the ring electrode. Mass analysis is achieved by making ion trajectories unstable in a mass-selective manner. (C) Triple quadrupole mass spectrometer, the ions of a particular
1.4.2.1 Quadrupole Mass Analyzer
The quadrupole is a mass filter, consisting of four rods to which direct current (DC) and a radio frequency (RF) alternate current (AC) are applied. Only ions carry a certain m/z can pass through, and reach the detector; while others are on unstable trajectories and fail to reach the detector. When these DC and RF voltages are increased, maintaining their ratios constant, ions with increasing value of m/z are recorded at the detector [36, 37]. Up to now, most experiments have been performed on triple quadrupole mass spectrometer that consists of three parts. Two mass separating quadrupoles was divided by a central quadruple whose function is to fragment the selected ion. Due to the presence of two independent quadrupoles, the triple quadrupole can be programmed for a variety of different scan modes, product ion scan, precursor ion scan, neutral loss scan and multiple reaction monitoring [38].
1.4.2.2 Ion Trap Mass Analyzer
The principle of three dimensional ions trap (3D ion trap) is a close relative of the quadrupole mass analyzer. Whereas a quadrupole has electric fields in two dimensions (x and y direction) and the ions move perpendicular to the field (z direction), the 3D ion trap has the electric field in all three dimensions, which can result in ions being trapped in the field. Unlike quadrupole, the spectrum is obtained by increasing the RF voltage that makes ions unstable and ejects for detection. For a MS/MS acquisition, all ions except the selected ion are ejected first, Subsequently, the remaining ion is fragmented and the product ions are analyzed [39, 40]. Ion trap is a robust, sensitive and relatively inexpensive instrument, which has successfully acquired much proteomics data in the literature. A disadvantage of ion trap is their relatively low mass accuracy due to the limited number of ions that can be accumulated, The space charge in ion trap distorts the accuracy of the mass measurement. Owing to the operating m/z are selected in a first quadrupole (Q1), fragmented in a collision cell (q2), and the fragment ions are separated in the last quadrupole (Q3). (D) TOF-TOF mass spectrometer, the ions of different m/z values have different velocities and therefore reach the detector at different times. It incorporates a collision cell between two TOF sections. Ions of one m/z are selected in the first TOF, fragmented in the collision cell, and the fragment ions are separated in the second TOF. (E) Fourier-transform ion-cyclotron resonance (FT-ICR) mass spectrometer, the ions oscillate around the magnetic field at frequencies that are related to their m/z. As ions oscillate near the top and bottom metal plates of the cubic trapping cell, they induce an alternating current that can be measured and then transferred to m/z. (F) Orbitrap mass spectrometer, the ions are trapped in its static electrostatic fields, in which the ions orbit around a central electrode and oscillate in axial direction, converting time domain signal into m/z like FT-ICR.
principle of the ion trap, the lower end of the fragment mass range cannot be detected (1/4 low mass cut off) [40].
1.4.2.3 Time-Of-Flight (TOF) Mass Analyzer
A TOF analyzer separates the ions based on velocity. It can be thought of as a race from the same starting point to the detector. Theoretically, all ions are formed at the same time and place in the ion source. Subsequently, the ions are accelerated to a fixed kinetic energy and travel down a flight tube. The low m/z ions achieve higher velocity than the high m/z ions. The spectrum is recorded by the impact of each ion on the detector. In fact, ion velocity is inversely related to the square root of m/z [41].
1.4.2.4 Fourier-Transform Ion-Cyclotron Resonance (FT-ICR) Mass Analyzer The FT-ICR mass spectrometer is also a trapping mass analyzer. It captures the ions under high vacuum in a high magnetic field. Once trapped, the ions oscillate with a cyclotron frequency that is inversely related to their m/z. The trapped ions are excited by an electric RF with a frequency in resonance with their cyclotron frequency. Although the ion oscillation ratio increases, its frequency is maintained, generating the image current for detection. The frequencies related to m/z can be calculated by a complex mathematical procedure (Fourier-transform, FT). As the frequencies can be measured precisely, the high-resolution and high-precision mass measurement is achieved under high magnetic field [42].
1.4.2.5 Orbitrap Mass Analyzer
The recent development of a novel Orbitrap mass spectrometry has made the exciting new areas of proteomic application possible. The Orbitrap was invented by Alexander Makarov in 1999 [43] and was reported as a tool for proteomics research in 2005 by Hu et al. [44]. In the Orbitrap analyzer, the ions are trapped and the orbit around a central spindle-like electrode and oscillate harmonically along its axis with a frequency characteristic of their m/z values, inducing an image current in the outer electrodes that is processed by Fourier-transform and generates the mass spectrum. The frequency of these harmonic oscillations is independent of the ion velocity and is inversely proportional to the square root of the m/z. The instrument is capable of mass resolution in excess of 100000 and mass accuracy of less than 2 ppm.
To summarize, MALDI is usually coupled to TOF mass analyzer, whereas ESI has mostly been coupled to ion traps, triple quadrupole, FT-ICR and Orbitrap mass analyzer. More recently, other new combinations of mass spectrometry are developed. For example, ESI quadrupole-TOF (Q-TOF) consists of ESI ion source coupled to the TOF analyzer [45]. MALDI TOF-TOF, MALDI ion source has been coupled to two types of TOF instruments. In the first, second TOF sections are separated by a collision cell [46]. These mass spectrometers have
high sensitivity, resolution and mass accuracy, and can be widely used in proteomic research.
1.4.3 Tandem Mass Spectrometry
Tandem mass spectrometry (MS/MS), as the name implies, involves two stages of MS. In the first stage, ions of a desired m/z are isolated (precursor ions). The isolated ions are increased the internal energy, and are induced to collide with an inert gas such as helium, argon or nitrogen, leading to dissociation. The resulting ions (product ions) are analyzed with the second stage of MS. MS/MS is a key technique for protein or peptide sequencing and PTMs analysis. Collision-induced dissociation (CID) has been the most widely used MS/MS technique in proteomics research. In this method, a particular gas phase peptide/protein ions is isolated and subsequently the energy is imparted by collisions with inert gas. The energy causes the peptide to break apart on the peptide backbone. Figure 1.6 shows how the peptide fragment and how the fragment ions are designated.
The most common and informative ions are generated by fragmentation at amide bond between amino acids. The resulting ions are called b-ions if the charge is localized on the N-terminal part of the peptide and y-ions if the charge is localized on the C-terminal part [47, 48]. The common proteomic experiment is performed with trypsin digestion .The resulting peptides have arginyl or lysyl residues at their C-terminus. In this case, the y-ion series are the predominant in the spectra. For an even more in-depth characterization, the fragment ions of peptide can be further fragmented. This is known as MS3 or more generally, MSn. It is worth to Figure 1.6. The types of peptide fragment ions observed in a MS/MS spectrum. (A) When the charge is localized on the N-terminus, the ion is classed as am, bm or cm. When the charge is localized on the C-terminus, the type of ion is classed as x(n-m), y(n-m) or z(n-m). (B) The fragment of peptide is induced by collisions with inert gas, and the bond breakage mainly occurs in the lowest energy pathway. That is cleavage of the amide bonds, which leads to form b- and y-ions series.
note that the internal fragmentation and neutral losses of H2O, NH3, and labile PTMs occur frequently during the CID process.
1.5 Data Analysis
As mentioned before, MALDI-TOF MS is still much used to identify protein by what is known as peptide mapping, also referred to as peptide mass fingerprinting (PMF), due to its simplicity, excellent mass accuracy, high resolution and sensitivity. In this method, protein is identified by matching a list of experimental peptide masses with the calculated list of all peptide masses of each entry in a protein database (SwissProt, NCBInr and etc.). Due to the fact that mass mapping requires an essentially purified target protein, the technique is commonly used in combination with protein separation technology using either SDS-PAGE or 2DE, respectively.
However, protein identification based solely on PMF should no longer be acceptable and must be complemented by MS/MS spectra. Because the MS/MS spectra contain structural information related to the sequence of the peptide, in addition to precursor ion mass information, these searches are generally more specific and discriminating. Several MS/MS spectra searching engines exist such as Sequest [49] and MASCOT [50]. The MS/MS spectra are collected as many as possible, and the results are searched by an algorithmic comparison via Sequest or MASCOT towards a protein database. These methods do not attempt to extract any sequence information at all from the MS/MS spectrum. Instead, the experimental fragment spectrum is matched against a calculated fragment spectrum for all peptides in the database.
A score is given to determine how close between MS/MS spectrum and the calculated peptide sequence. Sequest, a cross-correlation method, peptide sequences in the database are used to construct theoretical MS/MS spectra. The overlap or cross-correlation of these predicted MS/MS spectra with the experimental MS/MS spectra determines the best match. MASCOT, a probability based matching, the calculated fragment ion from peptide sequences in the database are compared with experimental MS/MS spectra. From this comparison, a score is calculated which reflects the statistical significance of the match between the MS/MS spectrum and the sequence contained in the database. In each of these methods, the identified peptides are compiled into a protein hit list, which is the output of a typical proteomic experiment. The protein identification relies on the matches with sequence database, and not all peptides resulting from the enzymatic digestion of a protein can be observed or correctly identified with MS analysis. This would reflect especially on the peptide with unexpected PTMs, and high-throughput proteomics is currently limited largely due to those species for which comprehensive sequence databases are available.
In brief, proteomics is the large-scale study of proteins, particularly their structures and functions. MS-based proteomics has become a formidable tool for the identification of proteins.
The limitation in dynamic range of MS analysis only allows for the proteins present at high
relative abundance to be preferentially identified, while information regarding the proteins present at low abundance in the complex mixtures is commonly not detected. Hence, the development of separation technology and the continued improvement of mass spectrometric methodology is crucial for identifying the low abundance proteins.
Chapter 2 - A High-Throughput Method for Phosphopeptide Enrichment of Spliceosomal Proteins and Its Applications
2.1 Summary
Reversible protein phosphorylation is a ubiquitous post-translational modification critical to many cellular processes. In this study, a simple, inexpensive and convenient titanium dioxide (TiO2) microspin column fabricated in a commercial pipette tip was developed for high-throughput enrichment of phosphopeptides in a crude mixture of spliceosomal proteins, which were digested by trypsin. The spliceosome is a protein-RNA complex which catalyses the excision of introns and ligation of exons of eukarytic pre-mRNAs. Our approach allows the enrichment of twenty-four samples at once. Careful comparison of our novel high-throughput method with the previously described manual TiO2 enrichment techniques showed similar result in terms of selectivity and sensitivity. Additionally, we evaluated and optimized the titania-based affinity enrichment for global profiling of phosphopeptides in total small nuclear ribonucleoproteins (snRNPs). We found that the use of RapiGest SF as detergent during digestion was more efficient than urea. The non-specific binding of non-phosphorylated peptides on TiO2 materials was reduced, but still maintained the high binding affinity of phosphopeptides without the need for an additional desalting step. Approximately 70 % of the enriched peptides were identified by mass spectrometry as being phosphorylated.
Furthermore, a complementary integrated analytical platform involving a combination of in-solution digestion, in-gel digestion from SDS-PAGE, TiO2 microspin columns, on-line nanoLC ESI-MS and off-line nanoLC MALDI-MS was employed to discover the maximum number of phosphorylation sites in the human spliceosomal proteins present in the crude mixture of nuclear snRNP particles. These strategies allow the complementary measurement of phosphopeptides. When compared with off-line nanoLC MALDI-MS/MS, online LC-ESI MS/MS turned out to be better for determining the exact location of the phosphorylation site. In total, 1381 phosphopeptides were identified in 390 proteins; of these, 640 sites were not previously described. The list of phosphopeptides was used to extract known and novel kinase motifs using the Motif-X algorithm. Finally, we showed three applications of this methodology for identifying phosphorylation sites and for studying protein-RNA crosslinks.
2.2 Introduction
A spliceosome is a complex of specialized ribonucleic acid (RNA) and protein subunits that removes introns from a transcribed precursor messenger ribonucleic acid (pre-mRNA) segment. The process is generally referred as splicing. Each spliceosome is composed of small nuclear RNA proteins, called snRNPs, and a range of non-snRNP associated protein factors. The snRNPs that make up the nuclear spliceosome are named U1, U2, U4, U5, and
U6, and participate in several RNA-RNA and RNA-protein interactions. The snRNPs display a broad variety of molecular sizes and chemical properties such as Arg-/Ser-rich tracts. They are ideally suited to establish methods for analysis of the changes in the phosphorylation during splicing process. Human spliceosome assembly intermediates have been observed include the E, A, B, B*, and C complex shown in Figure 2.1 [51]. The first recognition of pre-mRNA involves U1 snRNP binding to the 5' end splice site of the pre-mRNA and other non-snRNP associated factors to form the E complex. Subsequently, the U2 snRNP tightly associates with the branch point sequence (BPS) with the E complex to form A complex. The U4/U6.U5 tri-snRNP stably interacts to the assembling spliceosome to form complex B. Following several rearrangements in RNA-RNA and RNA-protein interactions, detaching the U1 and U4 snRNPs, give rise to the catalytically activated spliceosome (B* complex ) and then converts into C complex, in which the first of the two catalytic steps of splicing has occurred. After the second step, the spliceosome dissociates and the snRNPs are recycled for repeated rounds of spliceosome assembly.
PTMs of spliceosomal proteins play a crucial role in triggering conformational changes in protein-protein and RNA-protein interactions during the spliceosome assembly that are
Figure 2.1. The pathway of spliceosome assembly.
essential for its activity. For example, pre-mRNA splicing can be regulated both positively and negatively by reversible protein phosphorylation [52]. The phosphorylation of SR proteins are mediated by protein kinases / phosphatases and have been shown to be required for the formation of a catalytically active spliceosome [53-55]. Hence, an important focus of future will be to identify PTMs in spliceosomal proteins, and determine whether their modifications regulated during the splicing process.
Protein phosphorylation is one of the most common post-translational modifications and plays an important role in the regulation of a variety of cellular events, including signaling, cell differentiation, metabolism and apoptosis [56, 57]. Hence, the characterization of phosphorylation is a key issue in current proteomic research. Recently, MS-based techniques have been widely applied as powerful tools to characterize protein modifications, including phosphorylation, due to its speed, reliability, high sensitivity and capability for determining phosphorylation sites by MS/MS sequencing. However, large-scale phosphoproteomic analysis still remains a substantial challenge due to low abundance of phosphopeptides combined with large amounts of non-phosphorylated peptides which tend to suppress the ion signal of phosphorylated peptides in MS analysis [58-61]. Therefore, the highly specific separation and enrichment of phosphopeptides from proteolytic digest mixture becomes a critical step prior to MS analysis.
An effective method to resolving these problems is selective enrichment of phosphorylated peptides before MS analysis, like strong cation exchange (SCX) chromatography [62-66], strong anion exchange (SAX) chromatography [62, 64], immobilized metal ion affinity chromatography (IMAC) [67-74] and metal oxide affinity chromatography (MOAC) [75-88].
Among them, IMAC is commonly used and successfully coupled to various mass spectrometries. With this approach the phosphopeptides are captured by chelating interaction with metal ion such as Fe3+ or Ga3+ shown in Figure 2.2.
Figure 2.2. Phosphopeptide enrichment with immobilized metal ion affinity chromatography (IMAC). The binding between a phosphopeptide and IMAC resin is shown. The IMAC stationary phase is made by chelation of iminodiacetic acid (IDA) or nitrilotriacetic acid (NTA) with triply charged iron (Fe3+) or gallium (Ga3+) ions.
Enrichment and recovery strongly depend on the type of metal ion, column material, and loading/eluting procedures that are used. However, non-specific enrichment is unavoidable.
Acidic peptides tend to bind to the metal resins which complicate the MS detection of enriched phosphopeptides. Such non-specific binding can be reduced by prior methyl esterification of the acidic side chains of amino acid residues [89-92]. Nevertheless, incompleteness and side reactions of methyl esterification process can increase complexity of MS analysis and data interpretation, leading to decrease sensitivity.
Alternative to IMAC, more recently metal oxide affinity chromatography (MOAC) has been demonstrated to be effective material for the selective enrichment of phosphopeptides from proteolytic digests based on chelating interaction between phosphate functional groups and the surface of metal oxide particles. Metal oxides such as TiO2, ZrO2, Fe3O4 and Al2O3 are due to their high chemical stability ) [75-88]., tolerance over a broader pH range and physical robustness; hence, some non-volatile acidic additives, including 2,5-dihydroxybenzoic acid (DHB), phthalic acid (PA) and acidic buffers, can be employed to avoid non-specific binding [75, 88]. Among them, TiO2 has been widely used to selectively capture phosphopeptides shown in Figure 2.3.
Larsen et al. demonstrated that an addition of high quantity DHB in the loading and washing buffer could reduce non-specific bindings on TiO2 column [75]. However, although a diversity of techniques is available for phosphopeptides enrichment, the mapping of entire phosphoproteome is still a challenging task.
In this study, we invented a simple-to-fabricate, easy-to-use, economic and high efficiency TiO2
microspin column to enrich phosphopeptides. This design reduces the entire analyzed time in large scare analysis. Three commercially available articles for daily use - coffee filter, pipette tip and eppendorf tube - are utilized to fabricate microspin column. In addition, we integrated our microspin column with different proteomic technique to explore the phosphorylation sites in a crude mixture of nuclear snRNP particles, individual U snRNP, spliceosomal complexes and SR proteins.
Figure 2.3. Phosphopeptide enrichment with titanium dioxide (TiO2) bead. The binding between a phosphopeptide and TiO2 coated resin is shown
2.3 Experiment Sections 2.3.1 Materials
Iodoacetamide (IAA), -cyano-4-hydroxy-cinnamic acid (CHCA), 2,5-dihydroxybenzoic acid (DHB), phthalic acid (PA), trifluoroacetic acid (TFA), [Glu]-Fibrinopeptide B (Glu-Fib) and ammonium bicarbonate were obtained from Sigma-Aldrich (St. Louis, MO). Sequencing grade, modified trypsin was obtained from Promega (Madison, WI). Dithiothreitol (DTT), formic acid, ammonia solution, acetonitrile (ACN), and ethanol were obtained from Merck (Darmstadt, Germany). RapiGest SF was obtained from Waters Corporation (Manchester, UK). Titanium dioxide (TiO2) resins were obtained from GL Sciences Inc. (Tokyo, Japan).
2.3.2 A Crude Mixture of Nuclear snRNP Particles, Individual U snRNP, Spliceosomal Complexes and SR Proteins Purification
U1, U2, U5 snRNPs and U4/U6.U5 tri-snRNPs were isolated from HeLa nuclear extract by anti-m3G cap-directed immunoaffinity purification with the m3G-specific antibody H-20 to obtain a crude mixture of nuclear snRNP particles [93] and followed glycerol gradient centrifugation to separate the individual snRNPs as described previously [94, 95].
Spliceosomal A, B and C complexes were isolated from in vitro splicing reactions by the MS2 affinity-selection method [96, 97]. SR proteins were isolated by two salt precipitation steps, ammonium sulfate and magnesium chloride precipitation [98]. All protein complexes were purified by Prof. Reinhard Lührmann‘s Laboratory.
2.3.3 Ethanol Precipitation
The purified protein complexes were precipitated by adding 3 volumes of ethanol and 1/10 volume of 3 M sodium acetate, pH 5.3. The mixture was vortexed, incubated at -20 ℃ for 2 hours and then centrifuged 17000 g at 4 ℃ for 30 min. The supernatant was removed and the pellet was washed with 500 μl of 80 % ethanol and centrifuged as above. Discarded supernatant, the pellet was evaporated with a SpeedVac.
2.3.4 In-Solution Digestion
Precipitated proteins were dissolved with 20 μl of 1 % RapiGest™ SF in 25 mM ammonium bicarbonate, pH 8.5, sonicated for 15 min, reduced with 10 μl of 50 mM DTT at 37 ℃ for 1 hour, alkylated with 10 μl of 100mM IAA at 37 ℃ for 1 hour, diluted with 60 μl of 25 mM ammonium bicarbonate, and subsequently digested with trypsin (1:20 enzyme to substrate ratio) at 37 ℃, overnight. Tryptic peptides were acidified with 50 μl of 5 % TFA at 37 ℃ for 2 hours and followed centrifugation at 17000 g for 10 min. The resulting supernatant was transferred to another eppendorf and dried on a SpeedVac for further analysis. Another
procedure employed urea as a denaturation reagent instead of RapiGest™ SF. The ethanol precipitated proteins were dissolved with 20 μl of 6 M urea, sonicated, reduced, alkylated and digested as above and subsequently evaporated with a SpeedVac.
2.3.5 In-Gel Digestion
Precipitated proteins were separated by electrophoresis on a 4-12 % NuPAGE® Novex Bis-Tris Gels system (Invitrogen Corporation, Carlsbad, CA). The gel was stained with Coomassie Blue and cut equally into twenty slices. Each slice was reduced with 60 μl of 50 mM DTT at 37 ℃ for 1 hour, alkylated with 60 μl of 100 mM IAA at 37 ℃ for 1 hour, and subsequently digested with trypsin at 37 ℃, overnight (the enzyme to substrate ratio is 1:20) as described previously [99].
2.3.6 In-House TiO2 Microspin Column Fabrication
A small plug of coffee filter (1 mm x 1.5mm) was placed at the end of the tip by using a capillary tube shown in Figure. 2.4. The coffee filter serves only as a frit to retain the TiO2 resins within the commercial pipette tip. A length of approximately 3 mm of TiO2 materials was packed in the end of a pipette tip. Subsequently, the pipette tip was place into an open hole eppendorf for phosphopeptide enrichment.
2.3.7 Comparison of Different TiO2 Enrichment Procedure
Figure 2.4. Picture of a self‐made TiO
2microspin column.
Six different procedures for phosphopeptide enrichment were investigated shown in Figure 2.5.
In procedure A, tryptic peptides were dissolved with 60 μl of 200 mg 2,5-dihydroxybenzoic acid (DHB) in 1 ml of 80 % ACN, 5 % TFA and then loaded onto TiO2 microspin column. The column was washed three times with 60 μl of 200 mg DHB in 1 ml of 80 % ACN, 5 % TFA and five times with 60 μl of 80 % ACN, 5 % TFA. Bound peptides were eluted three times with 40 μl of 0.3 N NH4OH (pH ≥ 10.5), and subsequently evaporated NH4OH with a SpeedVac for further analysis. The speed of centrifugation for each step had to be controlled at 3000 rpm for 5 min.
Procedure B and C were similar to the procedure A with the exception that an additional desalted step with C18 (Nucleosil 100-5 C18) or NH2 (Nucleosil 100-5 NH2) microspin column was employed before TiO2 enrichment, respectively. Procedure D, E and F were similar to procedure A, B and C, respectively, with the exception that the loading and washing buffer were replaced with saturated phthalic acid (PA) in 80 % ACN, 5 %TFA.
2.3.8 NanoLC-ESI and -MALDI Mass Spectrometry Analysis.
The resulting peptides were first loaded at a flow rate of 10 μl/min onto an in-house packed C18 trap column (1.5 cm, 360 μm o.d., 150 μm i.d., Nucleosil 100-5 C18, Macherey-Nagel, GmbH & Co. KG). The retained peptides were then eluted and separated on an analytical C18 capillary column (30 cm, 360 μm o.d., 75 μm i.d., Nucleosil 100-5 C18) at a flow rate of 300 nL/min, with a gradient from 7.5 to 37.5 % ACN in 0.1 % formic acid for 60 min, 120 min or 240 min, using an Agilent 1100 nano-flow LC system (Agilent Technologies, Palo Alto, CA), coupling with Z-spray source or a robotic Probot micro fraction collector (LCPackings/Dionex, Sunnyvale, CA) for ESI-MS and -MS/MS (Waters/Micromass Q-Tof Ultima API mass spectrometer, Milford, MA) or MALDI-MS and -MS/MS (4800 MALDI-ToF/ToF, Applied Biosystems, Framingham, MA) analysis, respectively. For ESI-MS and -MS/MS data dependent acquisition, 1 s survey scans were run over the mass range m/z 350 to 1600. A maximum of three concurrent MS/MS acquisitions were triggered for 2+, 3+,
Figure 2.5. Optimization of the purification procedure using microspin TiO2 column.
and 4+ charged precursors detected at intensity above 15 counts, after 3s of acquisition, the system switched back to survey scan mode. For LC-MALDI analysis, The eluent was channelled to a nano-Tee where it was premixed with a matrix solution (10 mg/mL
-cyano-4-hydroxy-cinnamic acid, CHCA and 2.5 fmol/μl Glu-Fib in 80 % ACN, 0.1 % FA) delivered at 1.7 μl/min by a syringe pump and directly spotted onto 384-well MALDI sample plate in every 30 s. Subsequently, MALDI-MS and -MS/MS detected and sequenced peptides.
1000 and 5000 shots were accumulated in positive ion mode MS and MS/MS, respectively.
For collision-induced dissociation (CID) MS/MS operation, argon was used as collision gas and the indicated cell pressure was set up 5 x 10-7, with the potential difference between the source acceleration voltage and the collision cell set at 1 kV. The resolution of time ion selector for precursor ion was set at 200. MS spectra were acquired using Glu-Fib for internal calibration and MS/MS spectra were acquired using instrument default calibration.
2.3.9 Interpretation of Tandem Mass Spectra
All spectra were searched MASCOT server v2.2.0.6 against the NCBInr database limited to human with criteria-peptide mass tolerance, 50 ppm; MS/MS ion mass tolerance, 0.25 Da;
allowed up to three missed cleavage; variable modifications were considered phosphorylation of serine, threonine and tyrosine, methionine oxidation and cysteine carboxyamidomethylation.
The significant protein hits defined as protein score must be higher than 50 and if the protein score was between 20 and 50, we manually evaluated each MS/MS spectra. The threshold for individual peptide score must be higher than 20 and required bold red. All phosphorylated sited were examined manually by the presence of a 69 Da between fragment ions for phosphoserine and an 83 Da for phosphothreonine. Motif analysis was performed using motif-x [100] (http://motif-x.med.harvard.edu/) with significance set to 0.000001 and occurrence set to 20, using the human IPI as background. All amino acid frequency plots weblogos [101] (http://weblogo.berkeley.edu/) were created as frequency plots.
2.4 Result and Discussion 2.4.1 TiO2 Microspin Column
Recent reports have shown that TiO2 can be widely used to enrich phosphopeptides from peptide mixtures; however, the entire enriched procedure requires approximately 15 min for each sample [75, 102, 103]. Hence, we set out to establish a robust and fast (i.e.
semi-high-throughput) enrichment procedure for phosphopeptides derived from in-gel digestion and in-solution digestion. TiO2 materials are packed into a pipette tip that contains ordinary (coffee) filters as a frit to prevent leakage of TiO2 resins during the enrichment procedure (Figure 2.4). Hundreds of TiO2 microspin column were fabricated and tested during the period of this study, but we did not find any TiO2 resins slip out of the column. The system
has the further advantage that the device is small enough to fit into a 1.5 or 2 ml eppendorf tube. It can thus be easily placed into a benchtop laboratory centrifuge with a rotor for 1.5 (2 ml) eppendorf tubes. In this manner 24 samples (using a standard rotor) can be processed in parallel, the cost is reasonable and one-time usage eliminates the risk of carryover. In contrast, phosphopeptide enrichment with TiO2 microcolumns packed with gel loader tips, which are widely used in many labs, is much more time consuming due to manual handling, i.e. loading, washing and eluting procedure.
2.4.2 Sensitivity of the TiO2 Microspin Column
In initial experiments the specificity and sensitivity of our semi-automatic device for the enrichment of phosphopeptides was tested and compared with that of TiO2 microcolumn packed in gel loader tip [75]. We examined the power of TiO2 enrichment for phosphopeptides that derived from the arginine/serine-rich splicing factor SFRS1, SC35 and 9G8 separated by SDS-PAGE [98] as showed in Figure 2.6.
Fig 2.6A showed a direct MALDI peptide mass fingerprint of an aliquot of extracted tryptic peptides derived from arginine/serine-rich splicing factor SFRS1, SC35 and 9G8 without enrichment. The resulting spectrum was dominated by the signals of non-phosphorylated
Figure 2.6. MALDI mass spectra of tryptic peptides of arginine/serine-rich splicing factor SFRS1, SC35, 9G8. (A) Before enrichment. (B) Enrichment with the existing manual TiO2 microcolumn. (C) Enrichment with our TiO2 microspin column. The inset shows the SDS-PAGE gel image from purified SR proteins. The signals of phosphorylated peptides are marked with asterisk *.
peptides. After enrichment with TiO2 microcolumn packed in gel loader tip (Figure 2.6B) or with our TiO2 microspin column (Figure 2.6C, see Experiment Section 2.3.7, Procedure A in detail), the resulting spectra indicated the majority of non-phosphorylated peptides were removed, and the signals corresponding to the phosphopeptides were enhanced (marked with a star), which were confirmed by ESI-MS/MS (the peptide sequence with the phosphorylation site is listed within the spectrum, see also Appendix 1). In total, fourteen phosphopeptide signals were observed in both enrichment methods. One additional phosphopeptide signal at m/z 973.6 was found during enrichment with our TiO2 microspin column identified by our microspin column.
Figure 2.7. (A-B) Extracted ion chromatograms of peptide acetyl-SYGRPPPDVEGMTSLK at m/z 888.454 and phosphopeptide SKSPPKSPEEEGAVSS at m/z 888.378, respectively.
(C-D) MS/MS spectra of peptide acetyl-SYGRPPPDVEGMTSLK and phosphopeptide SKSPPKSPEEEGAVSS, respectively. The inset showed the m/z of precursor ion in the mass spectrum.
Strikingly, the abundant peak in the non-enriched spectrum at m/z 1775.8 (Figure 2.6A) was shown by MS/MS spectrum to be the N-terminal acetylated peptide SYGRPPPDVEGMTSLK of 9G8 protein (Figure 2.7B), whereas in both the enriched samples at m/z 1775.7 (Figure 2.6B and 2.6C), it was found to be the phosphopeptide SKSPPKSPEEEGAVSS (Figure 2.7D).
The mass difference between these two peaks was 0.152 Da. They were eluted separately from LC column at a retention time of 38.33 min and 22.23 min, respectively (Figure 2.7A and 2.7C). The specificity and sensitivity of our TiO2 microspin column is equivalent to existing manual microcolumn. Importantly, our TiO2 microspin column allows a significant higher sample throughput within a given time. In this manner phosphopeptides derived from proteins from in an entire gel lane that was cut into 23-24 slices (e.g. NUPAGE gels) can be enriched in parallel (see also below).
2.4.3 Optimizing Phosphopeptide Enrichment with TiO2 microspin column for In-Solution Digestion
We next made use of the high-throughput capabilities of the TiO2 microspin columns and systematically tested for maximum selectivity and sensitivity upon various in-solution digestion procedure, desalting steps and washing conditions that were most favorable to identify the maximum number of phosphopeptides. A crude mixture of total U snRNPs isolated from Hela nuclear extraction by immunoaffinity chromatography [93] was used as a test system.
Initial optimization of the procedure was evaluated by testing first different surfactant on the impact of TiO2 enrichment procedure. We first compared urea and RapiGest SF as denaturing agent (Figure 2.8.A-G and 2.8H-N, respectively). 10 μg of a crude mixture of total U snRNPs was denatured, reduced, alkylated and digested with trypsin over night. The resulting peptides were enriched as above (see Experiment Section 2.3.7 and Figure 2.5 in detail) and eluted peptides were analyzed by LC-ESI-MS/MS. Without enrichment only seven and nine phosphopeptides were identified in RapiGest SF and urea denatured sample, respectively (Figure 2.8A and 2.8H). In general, the total number of peptides derived from urea treated samples was higher as compared to RapiGest SF as the total number of peptides included those that are carbamylated at N-terminus, lysine, arginine and cysteine residues as shown in Figure 2.9 [104-106].
We found that digestion with RapiGest SF significantly increased the number of sequenced phosphopeptides (Figure 2.8B-G compared to 2.8I-N) and reduced the number of less non-phosphorylated peptides as compared to urea treatment. This can be explained several reasons: First, urea decomposes into ammonium cation and ammonium cyanate during the digestion process (Figure 2.9A). The ammonium cation elutes phosphopeptides from TiO2
resins and therefore decreases the number of phosphopeptides that can be identified.
Secondly, the lone pair electrons of nitrogen on the carbamyl group of carbamaylated peptides also binds to TiO2 resins. Third, RapiGest SF is an acid-labile surfactant. It hydrolyzes in acid solution posterior to in-solution enzymatic digestion to form sulfonic sodium salt in the sample shown in Figure 2.10 [107]. Carboxyl acid peptides in the sample that also bound to Figure 2.8. Selectivity and sensitivity of the different phosphopeptides enrichment strategies using TiO2 microspin column. Various surfactants , desalting steps, loading and washing buffers were compared to identify the maximum number of phosphopeptides and phosphoproteins by nanoLC-ESI-MS/MS as shown in A-N, respectively. (A-G) 10 μg total U snRNPs were digested in the presence of RapiGest SF. (H-N) instead of RapiGest SF with urea. (A) and (H) without enrichment. (B) and (I), procedure A, tryptic peptides were dissolved with 2,5-dihydroxybenzoic acid (DHB) buffer and subsequently loaded onto TiO2 microspin column; washed with DHB buffer and 5 % TFA in 80 % ACN and then eluted with NH4OH. (C) and (J), procedure B, an additional C18 desalting step was performed prior to TiO2 enrichment. (D) and (K), procedure C, instead of C18 with NH2
material. (E) and (L), procedure D, (F) and (M), procedure E, (G) and (M), procedure F, were the same as procedure A, B and C, respectively, with the exception that loading and washing buffer were replaced with saturated phthalic acid (PA). (see Experiment Section 2.3.7 and Figure 2.5 in detail). Specificity: Num. of phosphopeptides/Num. of identified peptides x 100 %.
TiO2 beads are displaced by the sulfonic anion but phosphopeptides were still retained, therefore, the specific for phosphopeptide in TiO2 enrichment is increased when using RapiGest SF [84, 102].
Since the surfactant deeply influenced the selectivity and sensitivity of TiO2 beads, we tested C18 and NH2 materials for removal of the surfactant prior to TiO2 to get more specificity and efficiency of enrichment. C18-based extraction method had been widely used in proteomics and other analyses for concentrating and desalting peptides [108, 109]. Only 44 and 31 phosphopeptides were identified by LC-MS/MS after desalting with C18 material and subsequent enrichment with TiO2 using RapiGest SF and urea, respectively (Figure 2.8C and 2.8J). Thus, the specificity of phosphopeptide enrichment was decreased from 68.4 % to 57.9 % in RapiGest SF experiment, and from 41.5 % to 20.5 % in the urea experiment. The missed phosphopeptides may be too hydrophilic or stuck in eppendorf and pipette tip during Figure 2.10. Decomposition of RapiGest SF in acid solution to form water immiscible compound and sulfonic sodium salt.
Figure 2.9. Formation of carbamylated proteins derived from urea. (A) A simple decomposition reaction of urea to form ammonia and cyanic acid (or isocyanic acid). (B) Carbamylation of proteins and peptides as isocyanic acid reacts with the N-terminus, lysine, arginine or cysteine residue.
